NSF Awards: 1657002
The Coding Science Internships project aims to create and research simulated internship curriculum units that confront barriers to broader participation in computer science and position coding as a tool for doing science. The curricular resources engage middle school students in code-to-learn experiences in which they deepen their science understanding while they develop computational thinking and core programming practices. By integrating coding within core science courses and providing teachers with structured, educative curriculum materials, the Internships are designed to expand participation in coding beyond self-selected populations and build teacher capacity for integrating computer science concepts and practices in science courses.
The two Coding Science Internships that we are developing and researching are designed to provide a motivating, student-centered learning experience in which students work collaboratively to address meaningful real-world problems. As an example, student intern teams use evidence and real-world data to program a scientific simulation to represent a coral reef ecosystem under threat. The purpose of the simulation is to communicate to stakeholders how various threats affect the health of a coral reef and how those threats may be mitigated. Students also gain first-hand experience with sequences, loops, and conditionals as they write code to program underwater robots to remove threats in variable conditions. Student teams work together to evaluate their code by testing executed code against expected outcomes and real-world data.
Kristana Textor
Wow, your video presents your idea so clearly and looks so polished! Can you tell us more about the resources you needed to produce it?
Ari Krakowski
Hi Kristiana,
Thanks! We have an amazing media content producer here at the Hall, Kat Quigley, that we worked with. To produce the video, we wrote a scripted outline and then worked with her to shoot and edit the video.
Best,
Ari
Paul Craig
It was fascinating to see the students start out working in isolation (them and their laptops) and then transition to "thinking out loud" with each other. How long did that transition take? Did you identify any barriers to collaboration or any approaches that helped to break down these barriers?
Ari Krakowski
Hi Paul,
We did several rounds of iterative pilot testing before having middle school teachers teach the unit in research trials. In our early pilot rounds, we had students working 1:1 on their own devices, which we soon saw was a barrier to student collaboration. We transitioned to a 1:2 pair programming device really helped foster student discourse and collaboration. Also, in our pilot revisions for the research trials, we developed student-facing collaboration norms, and a collaboration routine with roles (e.g. the driver and navigator roles used by code.org). And we provided some educative supports for teachers in the instructional guide and professional learning materials to help support them in facilitating student collaboration and discourse in the classroom. One other aspect that greatly affected student collaboration was the design of the student coding tasks. We found that tasks that had more than one possible strategy to meet specified criteria, or tasks in which students compared different coding strategies helped to open up student discourse and collaboration.
Best,
Ari
Lorien smyer
This is such a great idea!
Ari Krakowski
Thanks Lorien!
Gerad OShea
Research Director
Thanks for sharing your work!
I'm really intrigued by the idea of presenting the content as an internship for students to complete during their core science course. Can you share how this experience is framed for participating students and your thoughts about how the internship framing may contribute to student learning?
Best,
Gerad
Eric Greenwald
Director of Assessment and Analytics
Hi Gerad,
Thanks for the question! The simulated internship model is one we developed as part of a prior NSF grant (DRL #1417939). Students watch a brief "welcome to Futura" (the fictional company) video on day 1, presented as if from their Project Director that orients them to their intern role and outlines the problem they need to solve and the broad criteria for a successful solution. They also receive a "dossier" for the project, which they continually use as a resource to complete their work (for example in this unit, it includes background information about coral reefs). Throughout the unit, the teacher takes a "guide-on-the-side" role, coordinating the back and forth between the Project Director and students.
Each day, students also receive an email (within a custom-created workspace interface for the internship) outlining their tasks for the day (which is presented in terms of the overarching project they are working on). The internship fiction is largely maintained through students' daily interactions with thiis Futura Workspace platform and through continual nesting of student experiences within the framework of the internship work they are "assigned" by the fictional project director.
Students (and teachers) are generally onboard and motivated by the role they are playing. We've found that it allows a more authentic and relevant, and less 'school-y' way to engage with the concepts and practices that seems to be promoting a deeper, more active investment in the learning--teachers report and we've observed that students dig into the ideas beyond what might be necessary to 'get the right answer.' To further support this, we develop tasks (big and small) that can be approached in a variety of ways and enable a variety of acceptable solutions, which helps to democratize the knowledge/authority for knowledge, rather than positioning the teacher as the sole owner of the 'right answer. We also find that the internship structure creates new opportunities for authentic collaboration, as they are all in it together in trying to develop a working solution to the problem they, as interns, are tasked with.
Gerad OShea
Rebecca Grella, Ph.D.
Great presentation! I love the way you implement coding into environmental issues! WAY TO GO! I really would love to collaborate from the east coast!
Eric Greenwald
Director of Assessment and Analytics
Thanks! We'll keep you posted on opportunities for collaboration!
Ari Krakowski
Thanks Rebecca--we'd be happy to talk more!
Feng Liu
Researcher
Thanks for sharing this great work! I like the approach to learning how to program via coding to learn science content knowledge by solving real-world problems. I can see multiple potential outcomes such as student interest in coding and science subject, student knowledge in coding and science content, etc. Could you share more information about how you measure the impact (e.g., outcomes looked at, and how they are measured, the impact design approach)?
Eric Greenwald
Director of Assessment and Analytics
Thank you for your interest!
We have been piloting measures this year to capture science/CS learning as well as dispositional constructs pre/post:
Our computer programming/computational thinking measure builds on measurement development work by Chris Schunn's team at Pitt to create middle school friendly CT scales that draw from item sets in SRI's PACT work.
We'll be working this summer to evaluate validity evidence for each of the scales, revising as appropriate, and administering as a pre/post survey in broad research trials next year.
Feng Liu
Researcher
These are super helpful. Thanks for sharing these great resource, Eric!
Steven Bayless
Great video! The way you articulated and illustrated the concepts was spot on! Thanks!
Eric Greenwald
Director of Assessment and Analytics
Thanks, Steven!
Nancy McGowan
Instructional Math Coach
Your program is very practical and addresses computer science as it should-with a real world context. I can only imagine the drive your students have in working through researched issues knowing that they could/will have an impact on the well being of a coral reef and its inhabitants. Great job!
Ari Krakowski
Thanks so much Nancy!
Perla Myers
I really enjoyed watching the video (beautiful!) and learning about your project! I love the idea, especially how the computer science and the students' core science courses can be connected in such a meaningful way, and the internship approach. What role do the teachers play in the simulation? Do some of the teachers have an opportunity to work with each other before the students engage?
Ari Krakowski
Hi Perla, thanks for your questions!
We developed the virtual internship model as part of an NSF-funded project (DRL #1417939). In the unit, the students are interns in a (fictitious) company called Futura. They are introduced to the unit with a welcome video from a Project Director (Kai, featured here in the showcase video), and they also receive email messages introducing each lesson in a custom inbox interface. Teachers act as Internship Coordinators, helping students to develop understanding of the science and coding concepts and facilitating the work that students are doing in the coding environment. In this "guide on the side role," teachers have the support from the curriculum that enables any teacher, regardless of coding experience, to integrate the coding and computational thinking concepts with the science concepts they are more familiar with. We also provide professional learning and materials, as well as just-in-time information about pedagogical moves or conceptual unpacking. In the first round of research trials, teachers did not work together before they taught the unit, but that is definitely something we want to explore in the next round!
Daryl Pfeif
Hi Ari. What a thoughtful approach to solving the internship dilemma and what a great topic - visually rich, interesting. Without even seeing your evaluations I can only guess it's a much more empowering exercise for students to use their critical thinking and develop coding skills to solve a global problem rather than just hearing about it in the news.
Ari Krakowski
Thanks Daryl--it was really amazing to see how kids felt empowered by the coding context!
Karen Callahan
That looks neat. I was going to ask if this is it a mock internship, or are they really planting coral in reef scaffolds, but I think you just answered that in another post. On the other hand, if it interfaced to a code with drivers for machines, it could plant coral. Planting a tray of mixed seedlings might be easier to have something to take home, though. The interface looks cleaner and yet more code-like than Labview or Knime. It looked like the different corals have different properties, do they also have companion coral properties? Are these just known/given? If the actual properties were hidden, then students could use data science and many generated reefs to deduce good and bad neighbors, for example. I would say that school has limited time available, but I find that a lot of time for a lot of kids goes to Fortnight. When I was a kid Sim City was as popular as any of those games, but aside from very blatant correlations, there wasn't a convenient way to actually collate and analyze the data available in SimCity.
Ari Krakowski
Hi Karen,
Right, as you mentioned, it is a virtual internship, not a real one. So kids aren't transplanting real corals--however, there ARE real bots that are in the prototype stage that carry out some of the same restoration tasks (e.g. transplanting coral, removing invasive species). It's also true that different species of corals have different properties with respect to things like growth rate and reproduction. Since the curriculum unit is only ten days, there wasn't time to go into those species-level differences in depth, but there is information provided to the teacher to enable that extension if desired. Agree that providing data about real coral populations that kids draw on to design and test different restoration scenarios would be a really cool addition!
Best,
Ari
Tom Yeh
I've seen several projects over the years about teaching students to apply computer science real world environment problems. But most of those projects are limited to using computer science to measure or model the problems, such as detecting air pollution. What sets your project apart is that your project goes beyond just measuring but providing students a chance to actually do something positive to make the environment better (i.e., the coral restoration internship). Have you seen efforts taking your approach further to other environmental issues? For example, I can imagine a similar internship about programming a robot to plant trees to address the deforestation problem.
Along the same line of thought, I wonder whether negative examples could be developed, for example, an "evil" internship where a student is asked to program a poacher robot to harvest corals, cut down trees, or hunt animals. This could give students a balanced view of the impact of computer science and robotics, both positive and negative.
Ari Krakowski
Interesting! I could see this approach more broadly applied to using computational approaches to problem-solving. And I agree that it's important to consider that computational approaches can also be part of a problem, as well as affording solutions.
Thanks for your feedback Tom!
Further posting is closed as the showcase has ended.